## Memorials for the goners
- Mark Pilgrim
- happle tea
- abstruse goose
- 909sickle
- channelate

- Show title text: for each `img[title]`, copy the title out as sibling `p`
- Remove resizing: for each `img`, remove srcset,height,width attr, remove ComicPress resizing parameters in src attr
- Guessing bonus panel: for SMBC, make a copy of `img` with extra `_after` in `src` filename. It may happen to differ from real panel, but that's less load on SMBC servers

- [x] Show title text: for each `img[title]`, copy the title out as sibling `p`
- [x] Remove resizing: for each `img`, remove srcset,height,width attr, remove ComicPress resizing parameters in src attr

Inoreader remembers items per feed, i hope. It'd be nice to short `items` by some date, before letting inoreader fetch them.

let's start small with a feed fetcher, persisting to files, and a html-rewriter:
1. [https://www.smbc-comics.com/comic/rss](SMBC)
1. copy `img` with extra `after`

we're using `miniserve` to keep the VM flying, but data is generated
before build time with `cargo run --bin oneshot`, using one-time
fetched XML files.